speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

setup jamo

Browse Source
This commit is contained in:
游雁 2023-11-24 00:54:44 +08:00
parent afae93b43d
commit b5d3df75cf
2 changed files with 6 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
funasr/datasets/dataset_jsonl.py
View File

@ -4,8 +4,8 @@ import torch.distributed as dist
import numpy as np
import kaldiio
import librosa
import torchaudio
import time
def load_audio(audio_path: str, fs: int=16000):
audio = None
@ -17,7 +17,9 @@ def load_audio(audio_path: str, fs: int=16000):
if ".ark:" in audio_path:
audio = kaldiio.load_mat(audio_path)
else:
audio, fs = librosa.load(audio_path, sr=fs)
# audio, fs = librosa.load(audio_path, sr=fs)
audio, fs = torchaudio.load(audio_path)
audio = audio[0, :]
return audio
def extract_features(data, date_type: str="sound", frontend=None):

2
setup.py
View File

@ -14,7 +14,7 @@ requirements = {
"humanfriendly",
"scipy>=1.4.1",
"librosa",
# "jamo", # For kss
"jamo", # For kss
"PyYAML>=5.1.2",
# "soundfile>=0.12.1",
# "h5py>=3.1.0",